NeMiG is a bilingual (en/de) news #dataset on the topic of #migration, which can be used, among others, to conduct controlled experiments on #polarization by news #recommendersystems. https://arxiv.org/abs/2309.00550 Joint work in the #ReNewRS project w/
@dwsunima @fizise
@kitkarlsruhe
via HeikoPaulheim@twitter
#knowledgegraph #recsys

I wonder if anyone has bothered to look into the mathematics of #recommendersystems , because it is really embarrassing that papers with with 0000s of citations end up mapping to some very banal conventional #statistical models.
OTH, the latter have missed tremendous opportunities to entrench themselves (and create job opportunities for statisticians) by forgetting special cases that powered applications before the era of cheap computing that started in the 1980s.
It is #pagerank all over again

New paper out in #humancomputerinteraction journal: A sensemaking system for grouping and suggesting stories from multiple affective viewpoints in #museums 50 days free access: https://www.tandfonline.com/eprint/AHINMIJGM93A9EM4IQFU/full?target=10.1080/07370024.2023.2242355

#hci #recommendersystems #recsys #affectivecomputing #storytelling #culturalAI

@academicchatter
@cscw @digitalhumanities

https://tandfonline.com/doi/full/10.1080/07370024.2023.2242355

Recommendation systems are useful, but sometimes we just want to discover new things by ourselves. In this article I present 6 alternatives to "the algorithm".

https://blog.ddavo.me/posts/death-to-the-algorithm

#algorithm #youtube #radio #ethics #blog #recommendersystems #AI #datagovernance

Fancy working with me? The Faculty of Science and Engineering at the University of Wolverhampton offers, among others, a PhD studentship on Large Language Models for Academic Search. Please get in touch with me if you’re interested. Look for the LASER project at https://www.wlv.ac.uk/schools-and-institutes/faculty-of-science-and-engineering/research/phd-studentships/

#phd #ai #informationRetrieval #studentship #largeLanguageModels #chatgpt #chatgpt4 #bard #nlp #computationalLinguistics #search #academicSearch #recommendersystems #university #research

Now that the @Recsys paper deadlines are approaching, it's a good time to think about where to send that paper you planned to write but didn't quite finish on time 🙂 Consider the PERSPECTIVES workshop, which will be held in conjunction with #RecSys2023. Paper submission deadline in July #recsys #recommendersystems
https://perspectives-ws.github.io/2023/

Successful PhD defense of Malte Ostendorff at University of Göttingen today on "Aspect-based Document Similarity for Literature Recommender Systems". Great work! Congratulations to the new doctor!

Nice implementation for #arxiv papers:
https://recsys.ostendorff.org/recommendations?seed=DhOvSfFZEs&q=knowledge%20graph%20embeddings%20literals

@schubotz @DFKI #recommendersystems #PhD @fizise

Content moderation and #RecommenderSystems are the most widely deployed #MachineLearning based action taking systems. By action taking, I mean an algorithm that suggests a particular decision that impacts human actions. Using this view of recommender systems, we can consider how to increase the #diversity of online content by re-ranking recommended items to encourage creation of diverse content in the long term.

https://arxiv.org/abs/2302.04336

Similarity and Consistency in Algorithm-Guided Exploration
https://www.cesifo.org/DocDL/cesifo1_wp10188.pdf
…are more likely to follow an algorithm that is more exploitative than they are, possibly interpreting the algorithm’s relative consistency over time as a signal of expertise. …results suggest that the consistency of an algorithm’s recommendations over time is key to inducing people to follow algorithmic advice in exploration tasks.
#ExperimentalEcon #RecommenderSystems

Social media: what happens when AI takes over?

AI is about to make recommender algorithms a whole lot more effective, and potentially more dangerous, but it doesn't have to be that way.

Researchers @alasaarela and @lukethorburn are separately working on recommender algorithms that optimise for trust rather than attention and conflict. [Reg wall]

https://www.computing.co.uk/analysis/4074470/social-media-happens-ai-takes

#technews #ai #algorithms #socialmedia #recommendersystems #bridgingsystems

Delighted to share with you our following new article from the ReNewRS project: Divided by the Algorithm? The (Limited) Effects of Content-and Sentiment-Based News Recommendation on Affective, Ideological, and Perceived Polarization
Kudos to the authors Katharina Ludwig, Alexander Grote, Andreea Iana, Mehwish Alam, Heiko Paulheim, Christof Weinhardt, Philipp Müller

https://journals.sagepub.com/doi/pdf/10.1177/08944393221149290

#nlp #recommendersystems #socialsciences @fizise @unimannheim @KIT_Karlsruhe

In the latest SIGIR Forum issue, we discussed offline evaluation for RL-based #RecommenderSystems and noted that the most common evaluation protocol, i.e., next-item prediction, is unsuited to such approaches.

(with Thibaut Thonet, Jean-Michel Renders, @mdr )
Paper ➡️ https://sigir.org/wp-content/uploads/2023/01/p03.pdf

#RecSys #ReinforcementLearning

@clive That’s really interesting. (Someone better tell Meta. :)

I would argue though that there *is* an algorithm they might be doing better than other platforms, and that’s in classifying the video content. I could see that having a multiplier effect on any recommendation algorithm.

But I found it interesting reading this right after reading Cory Doctorow’s essay on social quitting; particularly the part about the willingness to throw in wildcards because they aren’t focused on high profile influencers.

What I wonder is if they can afford to continue doing that in the long run, or will the economics lead them down the same path as Instagram, where ads and influencers start to dominate, and the things that make them unique go away?

https://doctorow.medium.com/social-quitting-f049b33ad3f6

#recommendersystems #imagerecognition

Any prior work on the use of #clustering for #coldstart users in #recommendersystems

#machinelearning

Particularly interesting conclusion: it's not the business models that fuel #disinformation or #HateSpeech but ultimately it boils down to political opinions and decisions. #transparency of #RecommenderSystems helps but won't fix the problem. #PublicDiscourse

I had the pleasure of cofounding the interdisciplinary #CARLA workshop on #concept research: https://conceptresearch.github.io/CARLA/
We also published an edited volume on the topic (open access): https://link.springer.com/book/10.1007/978-3-030-69823-2

I recently transitioned to an industry position, where I work as machine learning engineer on #RecommenderSystems.

Switching to Mastodon for the obvious reasons (this weird guy, who bought twitter and who is posting and doing weird stuff)... Looking forward to reading about your insights!

I would like to know your take on this #recommendersystems #machinelearning problem. If you know of good papers in this please let me know.

Usually when we build a ranking system, we come up with multiple estimated probabilities of actions, like P(click), P(reply), P(react), E(time spent in channel | click) etc.
Then people often sort items based on `w_click * P(click) + w_react * P(react) + w_sth_else * P(sth_else)`.

I'm looking into recommender systems. What's currently the state of the art of recommending that deals gracefully with cold-start and can leverage item-level information in addition to collaborative information?

I found LightFM from 2015, but presumable a lot has changed since then?

#recsys #machinelearning #recommendersystems

TikTok's innovation isn't its recommendation algorithm but its design https://knightcolumbia.org/blog/tiktoks-secret-sauce

#TikTok #recommenderSystems #TheAlgorithm

@jiminy_kirket Shout out for your great presentation on #recommendersystems at #normconf! Really informative, really enjoyed it.